WO2020177215A1 - 一种岩土体材料数字图像的分割阈值确定方法 - Google Patents

一种岩土体材料数字图像的分割阈值确定方法 Download PDF

Info

Publication number
WO2020177215A1
WO2020177215A1 PCT/CN2019/086147 CN2019086147W WO2020177215A1 WO 2020177215 A1 WO2020177215 A1 WO 2020177215A1 CN 2019086147 W CN2019086147 W CN 2019086147W WO 2020177215 A1 WO2020177215 A1 WO 2020177215A1
Authority
WO
WIPO (PCT)
Prior art keywords
gray
segmentation threshold
rock
gray level
image
Prior art date
Application number
PCT/CN2019/086147
Other languages
English (en)
French (fr)
Inventor
刘江峰
曹栩楼
邵建富
黄炳香
王刚
胡大伟
陈亮
陈树亮
Original Assignee
中国矿业大学
徐州江恒能源科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中国矿业大学, 徐州江恒能源科技有限公司 filed Critical 中国矿业大学
Priority to US16/965,307 priority Critical patent/US20210201497A1/en
Publication of WO2020177215A1 publication Critical patent/WO2020177215A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10056—Microscopic image
    • G06T2207/10061—Microscopic image from scanning electron microscope
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30181—Earth observation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Definitions

  • the invention relates to the technical field of image segmentation, in particular to a method for determining a segmentation threshold of a digital image of rock and soil materials.
  • the most critical step is to binarize the image.
  • the main purpose of binarization is to distinguish the pore structure of the rock and soil material in the SEM image from the surface soil skeleton structure, and extract the pore structure.
  • the pore structure is the main factor that determines the permeability of the rock and soil material. Therefore, how to accurately extract the pore structure is the most critical technology in the binarization process in the digital image method.
  • the determination of the segmentation threshold of the image determines the accuracy of the binarization of pore structure extraction. Therefore, it is extremely important to determine the determination of the segmentation threshold of the image.
  • the beneficial effect is that the method is adopted to realize the scientific transformation of three-dimensional parameters to two-dimensional, eliminates the interference of human subjective factors, and makes the research of soil microstructure more accurate.
  • the disadvantage of this invention lies in the fact that the threshold value extracted by the Leica QWin image analysis software on the SEM photos in this invention has been artificially confirmed, but the artificial confirmation process involves multiple calculations, which is cumbersome and also Error-prone, which is not conducive to long-term use.
  • the present invention proposes a method for determining segmentation thresholds of geotechnical material digital images.
  • a method for determining a segmentation threshold of a digital image of a rock and soil material comprising the following steps:
  • S4 Determine the size of the segmentation threshold T according to the second derivative of the gray histogram curve and the value range of the segmentation threshold T.
  • the method further includes: reading the SEM image of the rock and soil material, and acquiring the gray level i of each pixel in the SEM image, The total pixel number n i corresponding to the gray level i of each pixel.
  • the gray histogram curve of the image acquired in the step S1 is specifically as follows:
  • i represents the gray level
  • N represents the total number of image pixels
  • n i represents the total number of pixels in the image with the gray level i
  • L represents the number of gray levels
  • step S2 determines the value range of the segmentation threshold T as follows:
  • Step S2.1 Determine the number of peaks in the grayscale histogram curve according to the grayscale histogram curve
  • Step S2.2 Determine the structure of the rock and soil material according to the number of peaks
  • Step S2.3 Obtain the gray level i max of the pixel corresponding to the peak
  • Step S2.4 Determine the value range of the segmentation threshold T according to the structure of the rock and soil material and the gray level i max of the pixel corresponding to the peak.
  • step S2.2 it is determined that the structure of the rock and soil material is specifically as follows:
  • the structure of the rock and soil material is a pore-containing structure
  • the gray-scale histogram curve has two peaks, the structure of the rock and soil material is a cracked structure.
  • step S3 before acquiring the second derivative of the grayscale histogram curve includes: acquiring the first derivative of the grayscale histogram curve, where the first derivative is:
  • i represents the gray level
  • n i represents the total number of pixels corresponding to the gray level i of each pixel in the image.
  • determining the segmentation threshold T of the pore structure is specifically as follows:
  • i max is the gray level of the pixel corresponding to the peak
  • i represents the gray level
  • n i represents the total number of pixels corresponding to the gray level i of each pixel in the image
  • SA4.3 Determine the maximum value a imax of the second derivative within the range of the segmentation threshold T according to the second derivative;
  • SA4.4 Determine the gray level i T of the pixel corresponding to the maximum value of the second derivative a imax , and the segmentation threshold T is:
  • i T is the gray level of the pixel corresponding to the maximum value of the second derivative a imax .
  • the determination of the segmentation threshold T of the fracture structure is specifically as follows:
  • i max1 is the gray level of the pixel corresponding to the first peak
  • i max2 is the gray level of the pixel corresponding to the second peak
  • n i represents the total number of pixels of all pixels with a gray level of i in the image, and i represents the gray level;
  • SB4.3 Determine the maximum value a imax of the second derivative within the value range of the segmentation threshold T according to the second derivative;
  • i T is the gray level of the pixel corresponding to the maximum value of the second derivative a imax .
  • the present invention first determines the value range of the segmentation threshold through the gray histogram curve of the SEM image of the rock and soil material to be tested, and then determines the specific size of the segmentation threshold through the second derivative of the gray histogram curve. Continuously narrowing the range until the accurate value is obtained can further ensure the accuracy of the segmentation threshold value;
  • the present invention performs analysis based on the SEM image of the rock and soil material to be tested, which can ensure that the segmentation threshold more satisfies the needs of digital image binarization of the rock and soil material to be tested, thereby accurately extracting the rock and soil from the digital image.
  • the pore or crack structure of the material is distinguished from the surface soil skeleton structure;
  • the present invention provides an accurate segmentation threshold for subsequent in-depth research of rock and soil materials based on digital images, and provides effective technical support for accurately extracting the pore or crack structure of rock and soil materials.
  • Figure 1 is a schematic flow diagram of the present invention
  • Figure 2 is an SEM image of bentonite
  • Figure 3 is a schematic diagram of the basic unit of bentonite particles
  • Figure 4 is a schematic cross-sectional view of the principle of SEM scanning of rock and soil materials
  • Figure 5 is a schematic diagram of the rock and soil structure corresponding to the gray histogram curve
  • Figure 6 is a binary graph of compacted bentonite with different segmentation thresholds
  • Figure 7 is the binarization extraction process of the SEM image as the segmentation threshold increases
  • Figure 8 is an SEM image of a coal sample after fracturing
  • Figure 9 is the gray histogram curve of coal sample after fracturing
  • Figure 10 is the gray histogram curve of bentonite
  • Figure 11 is the second derivative curve of the gray histogram curve of bentonite
  • Figure 12 is the final binary image of bentonite
  • Figure 13 is the second derivative curve of the gray histogram curve of the coal sample after fracturing
  • Figure 14 is the final binarized image of the coal sample after fracturing.
  • this embodiment provides a method for determining the segmentation threshold of digital images of rock and soil materials.
  • the segmentation threshold calculated by this method can effectively extract the pore structure or crack structure in the SEM image, and it also conforms to the geotechnical The actual distribution of pores and cracks in the bulk material. This provides a good technical support for the study of the meso-mechanism of rock and soil materials based on the digital image method.
  • Figure 4 is a scanning electron microscope that produces an image of the sample surface by scanning the surface of the sample with a focused electron beam. microscope.
  • the microstructure image of the rock and soil material surface obtained by scanning with SEM electron microscope is a grayscale image, in which the grayscale of the grayscale image ranges from 0 to 255, with a total of 256 values, and the color depth of each pixel in the SEM image Both represent a gray level.
  • the gray level of the pixel representing the pore or crack structure in the final SEM image is generally between 0 and 90; while the rock and soil material
  • the protruding particles on the surface soil skeleton structure are relatively close, so the pixels representing the soil skeleton structure on the surface of the rock and soil material in the final image generally show a gray level of 150-255; at the same time, the surface of the rock and soil material
  • the soil skeleton structure occupies the majority, specifically greater than 50%, and is in the same plane, so its gray level is generally 90 to 150, and the number of corresponding pixels is the largest.
  • the method for determining the segmentation threshold specifically includes the following steps:
  • Step S1 Read the SEM image of the rock and soil material to be tested through the MATALB code, and obtain the gray level i of each pixel in the SEM image and the total pixel number n i corresponding to the gray level i of each pixel.
  • Step S2 Obtain the gray histogram curve of the SEM image of the rock and soil material to be tested, the specific process is as follows:
  • Step S2.1 According to the gray level i of each pixel, the total number of pixels n i corresponding to the gray level i of each pixel and the following formula, obtain the gray histogram of the gray image of the rock and soil material to be tested Various points on the curve:
  • i represents the gray level
  • N represents the total number of image pixels
  • n i represents the total number of pixels in the image with the gray level i
  • L represents the number of gray levels
  • Step S2.2 According to each P(i) in step S2.1, fit the gray histogram curve of the gray image of the rock and soil material.
  • Step S3 Determine the value range of the segmentation threshold T according to the gray histogram curve.
  • the specific process is as follows:
  • Step S3.1 Determine the number of peaks in the gray histogram curve according to the gray histogram curve
  • Step S3.2 Determine the structure of the rock and soil material to be tested according to the number of peaks, specifically:
  • the structure of the rock and soil material to be tested is a porous structure
  • Step S3.3 Obtain the gray level i max of the pixel corresponding to the corresponding peak according to the number of peaks;
  • Step S3.4 Determine the value range of the segmentation threshold T according to the structure of the rock and soil material to be tested and the gray level i max of the pixel corresponding to the corresponding peak.
  • the determination of the value range of the segmentation threshold T is also not fixed, and is determined by the number of peaks corresponding to the structure.
  • Step S4 Obtain the second derivative of the gray histogram curve, the specific process is as follows:
  • Step S4.1 Obtain the first derivative of the gray histogram curve, the first derivative is:
  • i represents the gray level
  • n i represents the total number of pixels corresponding to the gray level i of each pixel in the image
  • Step S4.2 Obtain the second derivative of the gray histogram curve, the second derivative is:
  • i represents the gray level
  • n i represents the total number of pixels corresponding to the gray level i of each pixel in the image.
  • Step S5 Determine the size of the segmentation threshold T according to the second derivative of the gray histogram curve and the value range of the segmentation threshold T.
  • the specific process is as follows:
  • Step S5.1 According to each a i in step S4.2, fitting to obtain the second derivative curve;
  • Step S5.2 Determine the maximum value a imax within the range of the segmentation threshold T in the second derivative curve, and at the same time clarify the gray level i T of the pixel corresponding to the maximum value a imax , the segmentation threshold T is:
  • i T is the gray level of the pixel corresponding to the maximum value of the second derivative.
  • This embodiment provides a method for determining the segmentation threshold of a digital image of a rock and soil material.
  • bentonite is selected as the rock and soil material to be tested.
  • the method for determining the segmentation threshold of the bentonite digital image is as follows:
  • Step SA1 Refer to Figure 2, which is the SEM image of bentonite. Read the SEM image of bentonite through the MATALB code to obtain the gray level i of each pixel in the bentonite SEM image and the total gray level i of each pixel. The number of pixels n i .
  • Step SA2 Obtain the gray histogram curve of the bentonite SEM image, the specific process is as follows:
  • Step SA2.1 According to the gray level i of each pixel, the total number of pixels n i corresponding to the gray level i of each pixel, and the following formula, obtain each point on the gray histogram curve of the bentonite gray image :
  • i represents the gray level
  • N represents the total number of image pixels
  • n i represents the total number of pixels in the image with the gray level of i
  • L represents the number of types of gray levels.
  • Step SA2.2 According to each P(i) in step SA2.1, fit the gray histogram curve of the gray image of bentonite. Refer to Fig. 10, which is the gray-scale histogram curve of bentonite. From the gray-scale histogram curve of bentonite, one can intuitively see the proportion of the number of pixels of each gray level i in the bentonite SEM image.
  • Step SA3 Determine the value range of the segmentation threshold T according to the gray histogram curve. The specific process is as follows:
  • Step SA3.1 Determine the number of peaks in the gray histogram curve according to the gray histogram curve.
  • the structure of bentonite is Pore Structure.
  • the porosity of the rock and soil mass is less than 30%. Compared with the surface soil skeleton, the pore structure occupies a small part.
  • the porosity of the dry compacted bentonite is 20%-30% according to the mercury intrusion method and the core overburden porosity automatic test system.
  • the pore structure It should be around 25%.
  • the pore structure of bentonite is divided into several different levels. The pore structure of bentonite observed through SEM electronic microscope scanning belongs to the apparent pore structure, so the porosity obtained by digital image method should be less than 20%.
  • Step SA3.2 According to step SA3.1, there is only one peak in the gray histogram curve of bentonite. Therefore, in this embodiment, only the gray level i max of the pixel corresponding to one peak needs to be determined.
  • Step SA3.3 Determine the value range of the segmentation threshold T according to the structure of the bentonite and the gray level i max of the pixel corresponding to the corresponding peak. Because the structure of the bentonite is a pore structure, and its gray-scale histogram curve has only one peak, so The value range of the segmentation threshold T of the bentonite digital image is:
  • i max is the gray level of the pixel corresponding to the peak.
  • the structure of bentonite is pore structure. Since the surface soil skeleton accounts for the majority, the corresponding pixel at the peak must belong to the surface soil skeleton. It is worth noting that the porosity of dry compacted bentonite is less than 30% through the mercury intrusion method and the core overburden porosity and permeability automatic test system test. Therefore, the pixels representing the pore structure in the digital image of bentonite must also be less than 30%. The surface soil skeleton structure in bentonite exceeds 70%, so the pixels corresponding to the peak part of the gray histogram must represent the surface soil skeleton.
  • the gray level corresponding to the pore structure is generally between 0 and 90, and the remaining gray level between 90 and 225 corresponds to the surface soil skeleton structure, so the gray level of the pixel representing the pore structure is smaller than that of the surface soil.
  • the gray level of the skeleton pixel is generally between 0 and 90, and the remaining gray level between 90 and 225 corresponds to the surface soil skeleton structure, so the gray level of the pixel representing the pore structure is smaller than that of the surface soil.
  • the gray level of the skeleton pixel is generally between 0 and 90, and the remaining gray level between 90 and 225 corresponds to the surface soil skeleton structure, so the gray level of the pixel representing the pore structure is smaller than that of the surface soil.
  • Fig. 6 shows the changing process of the binary graph of compacted bentonite as the segmentation threshold T gradually increases from 0 to 255.
  • the black part in Figure 6 is the pore structure extracted according to different segmentation thresholds, and obviously some are very unreasonable. But through this change process, it can be observed that as the segmentation threshold T gradually increases, the porosity also gradually increases.
  • FIG. 7 shows the variation process of the binarization image with the segmentation threshold T.
  • the increase in the porosity of the binarization image is actually the increase in the number of black pixels in the figure as the segmentation threshold T increases.
  • the segmentation threshold T reaches the gray level of the surface soil skeleton from the gray level of the pore structure, since the surface soil skeleton structure accounts for the majority, the total number of pixels n i corresponding to the gray level i of each pixel in the image will exist A mutation, therefore, find the gray level i corresponding to this point from the pore structure to the surface soil skeleton structure mutation, that is, determine the size of the segmentation threshold T.
  • Step SA4 Obtain the second derivative of the gray histogram curve, the specific process is as follows:
  • Step SA4.1 Obtain the first derivative of the gray histogram curve, the first derivative is:
  • i represents the gray level
  • n i represents the total number of pixels corresponding to the gray level i of each pixel in the image.
  • Step SA4.2 Obtain the second derivative of the gray histogram curve, the second derivative is:
  • i represents the gray level
  • n i represents the total number of pixels corresponding to the gray level i of each pixel in the image.
  • Step SA5 Determine the size of the segmentation threshold T according to the second derivative a i of the gray histogram curve and the value range of the segmentation threshold T.
  • the specific process is as follows:
  • Step SA5.1 According to each a i in step SA4.2, the second derivative curve is obtained by fitting, referring to FIG. 11, which is the second derivative curve of the gray histogram curve of bentonite.
  • Step SA5.2 Determine the maximum value a imax in the second derivative curve within the range of the segmentation threshold T, and determine the gray level i T of the pixel corresponding to the maximum value a imax .
  • the segmentation threshold T is:
  • i T is the gray level of the pixel corresponding to the maximum value of the second derivative.
  • This embodiment provides a method for determining the segmentation threshold of a digital image of a rock and soil mass material.
  • the rock and soil mass material to be tested selects a coal sample after fracturing.
  • the method for determining the segmentation threshold of the digital image of the coal sample after fracturing is as follows:
  • Step SB1 Refer to Figure 8.
  • Figure 8 is the SEM image of the coal sample after fracturing. Read the SEM image of the coal sample after fracturing through the MATALB code to obtain the gray level i of each pixel in the SEM image of the coal sample after fracturing. And the total pixel number n i corresponding to the gray level i of each pixel.
  • Step SB2 Obtain the gray histogram curve of the SEM image of the coal sample after fracturing.
  • the specific process is as follows:
  • Step SB2.1 According to the gray level i of each pixel, the total number of pixels n i corresponding to the gray level i of each pixel, and the following formula, obtain the gray histogram curve of the gray image of the coal sample after fracturing Various points on:
  • i represents the gray level
  • N represents the total number of image pixels
  • n i represents the total number of pixels in the image with the gray level of i
  • L represents the number of types of gray levels.
  • Step SB2.2 According to each P(i) in step SB2.1, fit the gray histogram curve of the gray image of the coal sample after fracturing. Refer to Figure 9, which is the gray histogram curve of the SEM image of the coal sample after fracturing.
  • Step SB3 Determine the value range of the segmentation threshold T according to the gray histogram curve. The specific process is as follows:
  • Step SB3.1 Determine the number of peaks in the gray histogram curve according to the gray histogram curve.
  • the gray histogram curve referring to Figure 9, it can be seen that there are two peaks in the gray histogram curve, so after fracturing The structure of the coal sample is a cracked structure.
  • Step SB3.2 According to step SB3.1, there are two peaks in the gray histogram curve of the coal sample after fracturing. Therefore, in this embodiment, it is necessary to determine the gray levels i max1 of the pixels corresponding to the two peaks respectively. And i max2 , where i max1 is the gray level of the pixel corresponding to the first peak, and i max2 is the gray level of the pixel corresponding to the second peak.
  • Step SB3.3 Determine the value range of the segmentation threshold T according to the structure of the coal sample after fracturing and the gray levels i max1 and i max2 of the pixels corresponding to the corresponding peaks, because the structure of the coal sample after fracturing is a pore structure, and The gray histogram curve has two peaks, so the range of the segmentation threshold T of the digital image of the coal sample after fracturing is:
  • i max1 is the gray level of the pixel corresponding to the first peak
  • i max2 is the gray level of the pixel corresponding to the second peak.
  • the gray histogram curves of rock and soil materials with cracked structures are all bimodal curves, so the structure of the coal sample after fracturing is rock and soil with cracked structures, and the gray level i increases from small to large.
  • the pixel at the first peak in the image represents the pixel at the crack
  • the pixel at the second peak represents the pixel at the surface skeleton, so the segmentation threshold T of the rock and soil material with crack structure Between two peaks.
  • Step SB4 Obtain the second derivative of the gray histogram curve, the specific process is as follows:
  • Step SB4.1 Obtain the first derivative of the gray histogram curve, the first derivative is:
  • i represents the gray level
  • n i represents the total number of pixels corresponding to the gray level i of each pixel in the image.
  • Step SB4.2 Obtain the second derivative of the gray histogram curve, the second derivative is:
  • i represents the gray level
  • n i represents the total number of pixels corresponding to the gray level i of each pixel in the image.
  • Step SB5 Determine the size of the segmentation threshold T according to the second derivative a i of the gray histogram curve and the value range of the segmentation threshold T.
  • the specific process is as follows:
  • Step SB5.1 Fit and obtain the second derivative curve according to each a i in step SB4.2, refer to Fig. 13, which is the second derivative curve of the gray histogram curve of the coal sample after fracturing.
  • Step SB5.2 Determine the maximum value a imax in the second derivative curve within the range of the segmentation threshold T, and determine the gray level i T of the pixel corresponding to the maximum value a imax .
  • the segmentation threshold T is:
  • i T is the gray level of the pixel corresponding to the maximum value of the second derivative.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)
  • Analysing Materials By The Use Of Radiation (AREA)

Abstract

本发明公开了一种岩土体材料数字图像的分割阈值确定方法，所述方法包括如下步骤：S1：获取岩土体材料的SEM图像的灰度直方图曲线；S2：根据所述灰度直方图曲线，确定分割阈值T的取值范围；S3：获取所述灰度直方图曲线的二阶导数；S4：根据所述灰度直方图曲线的二阶导数和分割阈值T的取值范围，确定所述分割阈值T的大小。本发明能够快速准确的确定岩土体材料的数字图像的分割阈值，准确地从数字图像中将岩土体材料孔隙或裂缝结构与表面土骨架结构区分开来，为后续基于数字图像的岩土体材料的深入研究提供了准确的分割阈值，为准确的提取孔隙或裂缝结构，提供了有效的技术支持。

Description

一种岩土体材料数字图像的分割阈值确定方法 技术领域
本发明涉及图像分割技术领域，尤其涉及一种岩土体材料数字图像的分割阈值确定方法。
背景技术
随着微、细观成像技术的发展，研究人员可以利用该技术对岩土体的孔隙结构形态特征进行直接观测。对于SEM扫描电镜技术，由于其较低的测试成本和良好的成像效果，被逐渐应用于各个领域。
使用SEM扫描电镜技术对各种岩土体材料的表面微观孔隙结构进行直接观测，通过定量化的分析，对岩土体材料渗透率进行预测。在对SEM图像定量化分析的过程中，最关键的一步就是对图像进行二值化。二值化的主要目的是将SEM图像中的岩土体材料的孔隙结构与表面土骨架结构区分开来，提取出其中的孔隙结构，而孔隙结构是决定岩土体材料渗透率大小的主要因素，因此如何准确的提取孔隙结构就是数字图像法中二值化过程最为关键的技术。而二值化处理过程中，图像的分割阈值确定决定着孔隙结构提取二值化的准确性，因此，确定图像的分割阈值确定显得极为重要。
近几十年来，国内外众多学者已经提出了大量的图像二值化分割阈值确定计算方法。但由于数字图像的复杂性，以及图像分割问题依赖于具体应用领域的具体情况，因此，直到现在为止没有一种通用的二值化分割阈值确定算法。
中国专利公布号：CN 102841220 A；公布日：2012年12月26日，公开了一种基于孔隙率的黏土扫描电镜照片图像分割方法，包括步骤有：采用土力学常规实验方法测得黏土试样的干密度，算出干密度的三分之二次幂；实验室制备黏土试样的微结构样品；采用ZD-A3冻干仪将制作的样品冻干并抽真空，真空度达到到10 -6Pa时进行扫描电镜，观察样品的微结构并获得扫描照片；采用图像分析软件提取两个阈值对应的二值化数据；对比两个阈值条件下抽取到颗粒体和孔隙率的参数特征，算出孔隙体面积比；当孔隙体面积比等于干密度的三分之二次幂时，停止逼近，此时的阈值即为与所测干密度相对应的图像分割阈值确定。有益效果是采用该方法实现了三维参数向二维化的科学转变，排除了人为主观因数的干扰，使得土体微结构的研究更加准确。但是该发明的不足之处在于：该发明中Leica QWin图像分析软件对SEM照片提取出的阈值，虽然进行了人为确认，但是在进行人为确认的过程中，涉及多次运算，过程繁琐的同时还易出错，从而不利于长期使用。
发明内容
发明目的：针对现有岩土体材料数字图像二值化中分割阈值的确定过程复杂易出错 的问题，本发明提出一种岩土体材料数字图像的分割阈值确定方法。
技术方案：为实现本发明的目的，本发明所采用的技术方案是：
一种岩土体材料数字图像的分割阈值确定方法，所述方法包括如下步骤：
S1：获取岩土体材料的SEM图像的灰度直方图曲线；
S2：根据所述灰度直方图曲线，确定分割阈值T的取值范围；
S3：获取所述灰度直方图曲线的二阶导数；
S4：根据所述灰度直方图曲线的二阶导数和分割阈值T的取值范围，确定所述分割阈值T的大小。
更进一步地，所述步骤S1获取图像的灰度直方图曲线之前还包括：读取所述岩土体材料的SEM图像，获取所述SEM图像中每个像素的灰度级i、所述每个像素的灰度级i所对应的总像素数量n i。
更进一步地，所述步骤S1获取图像的灰度直方图曲线具体如下：
S1.1：确定所述岩土体材料灰度图像中每个像素的灰度级i；
S1.2：根据如下公式，获取所述岩土体材料灰度图像的灰度直方图曲线上各个点：
Figure PCTCN2019086147-appb-000001
其中，i表示灰度级，N表示图像像素总数，n i表示图像中所有灰度级为i的像素的总像素个数，L表示灰度级的种类数；
S1.3：根据所述P(i)，拟合获取岩土体材料灰度图像的灰度直方图曲线。
更进一步地，所述步骤S2确定分割阈值T的取值范围具体如下：
步骤S2.1：根据所述灰度直方图曲线，确定所述灰度直方图曲线中的峰值数目；
步骤S2.2：根据所述峰值数目，确定所述岩土体材料的结构；
步骤S2.3：获取所述峰值对应像素的灰度级i max；
步骤S2.4：根据所述岩土体材料的结构和峰值对应像素的灰度级i max，确定所述分割阈值T的取值范围。
更进一步地，步骤S2.2确定所述岩土体材料的结构具体如下：
如果所述灰度直方图曲线只有一个峰值，所述岩土体材料的结构为含孔隙结构；
如果所述灰度直方图曲线有两个峰值，所述岩土体材料的结构为含裂缝结构。
更进一步地，步骤S3获取所述灰度直方图曲线的二阶导数之前包括：获取所述灰度直方图曲线的一阶导数，所述一阶导数为:
Figure PCTCN2019086147-appb-000002
其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量。
更进一步地，确定所述孔隙结构的分割阈值T具体如下：
SA4.1：所述灰度直方图曲线只有一个峰值，所述分割阈值T的取值范围为：
T<i max
其中，i max为峰值对应像素的灰度级；
SA4.2：获取所述灰度直方图曲线的二阶导数，所述二阶导数为：
Figure PCTCN2019086147-appb-000003
其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量；
SA4.3：根据所述二阶导数，确定在分割阈值T的取值范围内二阶导数的最大值a imax；
SA4.4：确定所述二阶导数最大值a imax所对应像素的灰度级i T,所述分割阈值T为：
T＝i T
其中，i T为二阶导数最大值a imax所对应像素的灰度级。
更进一步地，确定所述裂缝结构的分割阈值T具体如下：
SB4.1：所述灰度直方图曲线有两个峰值，所述分割阈值T的取值范围为：
i max1<T<i max2
其中，i max1为第一个峰值对应像素的灰度级，i max2为第二个峰值对应像素的灰度级；
SB4.2：获取所述灰度直方图曲线的二阶导数，所述二阶导数为：
Figure PCTCN2019086147-appb-000004
其中，n i表示图像中所有灰度级为i的像素的总像素个数，i表示灰度级；
SB4.3：根据所述二阶导数，确定在分割阈值T的取值范围内二阶导数的最大值a imax；
SB4.4：确定所述二阶导数最大值a imax所对应像素的灰度级i T,所述分割阈值T为：
T＝i T
其中，i T为二阶导数最大值a imax所对应像素的灰度级。
有益效果：与现有技术相比，本发明的技术方案具有以下有益技术效果：
(1)本发明先是通过待测岩土体材料的SEM图像的灰度直方图曲线，确定分割阈值的取值范围，再通过灰度直方图曲线的二阶导数确定分割阈值的具体大小，通过不断的缩小范围直至精确取值，能够更进一步地保证分割阈值取值的准确性；
(2)本发明根据待测岩土体材料自身的SEM图像进行分析，能够确保分割阈值更加满足待测岩土体材料数字图像二值化的需求，从而准确的从数字图像中将岩土体材料的孔隙或裂缝结构与表面土骨架结构区分开来；
(3)本发明为后续基于数字图像的岩土体材料的深入研究提供了准确的分割阈值，为准确的提取岩土体材料的孔隙或裂缝结构，提供了有效的技术支持。
附图说明
图1是本发明的流程示意图；
图2是膨润土SEM图像；
图3是膨润土颗粒基本单元的示意图；
图4是SEM扫描岩土体材料原理剖面示意图；
图5是灰度直方图曲线对应岩土体结构示意图；
图6是不同分割阈值压实膨润土二值图；
图7是SEM图像随分割阈值增大的二值化提取过程；
图8是压裂后煤样SEM图像；
图9是压裂后煤样灰度直方图曲线；
图10是膨润土的灰度直方图曲线；
图11是膨润土的灰度直方图曲线二阶导数曲线；
图12是膨润土的最终二值化图像；
图13是压裂后的煤样的灰度直方图曲线二阶导数曲线；
图14是压裂后的煤样的最终二值化图像。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述。其中，所描述的实施例是本发明一部分实施例，而不是全部的实施例。因此，以下对在附图中提供的本发明的实施例的详细描述并非旨在限制要求保护的本发明的范围，而是仅仅表示本发明的选定实施例。
参考图1，本实施例提供了一种岩土体材料数字图像的分割阈值确定方法，本方法计算得到的分割阈值能够有效地提取出SEM图像中的孔隙结构或裂缝结构，同时也符合岩土体材料孔隙及裂缝的实际分布情况。从而为基于数字图像法的岩土体材料细观机理研究提供了良好的技术支持。
参考图4，图4为扫描电子显微镜通过用聚焦电子束扫描样品的表面来产生样品表面的图像，其中扫描电子显微镜英文为：Scanning Electron Microscope，可以缩写为SEM，简称扫描电镜，是一种电子显微镜。
使用SEM电镜扫描得到的岩土体材料表面微观结构图像为灰度图像，其中灰度图像的灰度级从0到255，总共有256种取值，同时SEM图像中每个像素点颜色的深浅都代表着一个灰度级。
由于岩土体材料的孔隙结构或裂缝结构距离电子束发射口较远，因此最终的SEM图像中代表孔隙或裂隙结构的像素点呈现出的灰度级一般在0～90；而岩土体材料的表面土骨架结构上的凸起颗粒距离较近，因此最终图像中代表岩土体材料表面土骨架结构的像素点呈现出的灰度级一般在150～255；同时其中岩土体材料的表面土骨架结构占大多 数，具体大于50％，且处于同一平面中，因此其灰度级一般在90～150，对应像素数量最多。
在本实施例中，具体地讲，本分割阈值确定方法具体包括如下步骤：
步骤S1：通过MATALB代码读取待测岩土体材料的SEM图像，获取SEM图像中每个像素的灰度级i以及每个像素的灰度级i所对应的总像素数量n i。
步骤S2：获取待测岩土体材料的SEM图像的灰度直方图曲线，具体过程如下：
步骤S2.1：根据每个像素的灰度级i、每个像素的灰度级i所对应的总像素数量n i以及如下公式，获取待测岩土体材料灰度图像的灰度直方图曲线上的各个点：
Figure PCTCN2019086147-appb-000005
其中，i表示灰度级，N表示图像像素总数，n i表示图像中所有灰度级为i的像素的总像素个数，L表示灰度级的种类数；
步骤S2.2：根据步骤S2.1中的各个P(i)，拟合获取岩土体材料灰度图像的灰度直方图曲线。
步骤S3：根据灰度直方图曲线，确定分割阈值T的取值范围，具体过程如下：
步骤S3.1：根据灰度直方图曲线，确定灰度直方图曲线中的峰值数目；
步骤S3.2：根据峰值数目，确定待测岩土体材料的结构，具体为：
如果灰度直方图曲线中只有一个峰值，待测岩土体材料的结构为含孔隙结构；
如果灰度直方图曲线中有两个峰值，待测岩土体材料的结构为含裂缝结构；
步骤S3.3：根据峰值数目，获取相应峰值对应像素的灰度级i max；
步骤S3.4：根据待测岩土体材料的结构、相应峰值对应像素的灰度级i max，确定分割阈值T的取值范围。
值得注意的是，由于待测岩土体材料结构的不确定，分割阈值T的取值范围的确定也是不固定的，由结构对应峰值的数目决定。
步骤S4：获取灰度直方图曲线的二阶导数，具体过程如下：
步骤S4.1：获取灰度直方图曲线的一阶导数，一阶导数为:
Figure PCTCN2019086147-appb-000006
其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量；
步骤S4.2：获取灰度直方图曲线的二阶导数，二阶导数为：
Figure PCTCN2019086147-appb-000007
其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量。
步骤S5：根据灰度直方图曲线的二阶导数和分割阈值T的取值范围，确定分割阈 值T的大小，具体过程如下：
步骤S5.1：根据步骤S4.2中的各个a i，拟合获取二阶导数曲线；
步骤S5.2：确定在二阶导数曲线中，分割阈值T取值范围内的最大值a imax，同时明确最大值a imax所对应像素的灰度级i T,分割阈值T为：
T＝i T
其中，i T为二阶导数最大值所对应像素的灰度级。
实施例1 含孔隙结构的岩土体材料
本实施例提供了一种岩土体材料数字图像的分割阈值确定方法，在本实施例中，具体地讲，待测岩土体材料选择膨润土。其中膨润土数字图像的分割阈值确定方法，具体如下：
步骤SA1：参考图2，图2为膨润土的SEM图像，通过MATALB代码读取膨润土SEM图像，获取膨润土SEM图像中每个像素的灰度级i以及每个像素的灰度级i所对应的总像素数量n i。
步骤SA2：获取膨润土SEM图像的灰度直方图曲线，具体过程如下：
步骤SA2.1：根据每个像素的灰度级i、每个像素的灰度级i所对应的总像素数量n i以及如下公式，获取膨润土灰度图像的灰度直方图曲线上的各个点：
Figure PCTCN2019086147-appb-000008
其中，i表示灰度级，N表示图像像素总数，n i表示图像中所有灰度级为i的像素的总像素个数，L表示灰度级的种类数。
步骤SA2.2：根据步骤SA2.1中的各个P(i)，拟合获取膨润土灰度图像的灰度直方图曲线。参考图10，图10为膨润土的灰度直方图曲线，从膨润土的灰度直方图曲线中，能够直观的看到膨润土SEM图像中每个灰度级i的像素数量的占比情况。
步骤SA3：根据灰度直方图曲线，确定分割阈值T的取值范围，具体过程如下：
步骤SA3.1：根据灰度直方图曲线，确定灰度直方图曲线中的峰值数目，在本实施例中，参考图10可知，灰度直方图曲线中的峰值只有一个，因此膨润土的结构为孔隙结构。
根据压汞法及岩心覆压孔渗自动测试系统测试得到岩土体的孔隙率小于30％，其与表面土骨架相比来说，孔隙结构占小部分。在本实施例中，具体地讲，根据压汞法及岩心覆压孔渗自动测试系统测试得到干燥压实膨润土的孔隙率为20％～30％，对于膨润土SEM图像来说，其中的孔隙结构应该在25％左右。参考图3可知，膨润土的孔隙结构又分有几个不同的层次，通过SEM电子镜扫描观察到的膨润土孔隙结构属于表观孔隙结构，故通过数字图像法得到的孔隙率应该小于20％。
步骤SA3.2：根据步骤SA3.1可知，膨润土的灰度直方图曲线中只有一个峰值，因此在本实施例中，只需要确定一个峰值对应像素的灰度级i max。
步骤SA3.3：根据膨润土的结构、相应峰值对应像素的灰度级i max，确定分割阈值T的取值范围，因为膨润土的结构为孔隙结构，且其灰度直方图曲线只有一个峰值，故膨润土数字图像的分割阈值T的取值范围为：
T<i max
其中，i max为峰值对应像素的灰度级。
参考图5，由于所有含孔隙结构的岩土体材料的灰度直方图都是单峰曲线，所以膨润土的结构为孔隙结构。其中由于表面土骨架占大多数，因此峰值处对应像素一定属于表面土骨架。值得注意的是，通过压汞法及岩心覆压孔渗自动测试系统测试得到干燥压实膨润土的孔隙率小于30％，因此，膨润土的数字图像中代表孔隙结构的像素也肯定小于30％，故膨润土中的表面土骨架结构超过70％，因此灰度直方图中峰值部分对应像素肯定是代表表面土骨架。显然地，孔隙结构对应的灰度级一般在0～90，而剩下的在90～225的灰度级对应的则为表面土骨架结构，故代表孔隙结构像素的灰度级小于代表表面土骨架像素的灰度级。
参考图6，图6为随着分割阈值T从0逐渐增大到255时，压实膨润土二值图的变化过程。其中，图6中黑色部分就是根据不同的分割阈值提取出的孔隙结构，显然有一些是非常不合理的。但是通过这个变化过程，可以观察到随着分割阈值T的逐渐增大，孔隙率也在逐渐增大。
参考图7，图7为二值化图随着分割阈值T的变化过程，其中二值化图像孔隙率的增大实际上就是图中黑色像素数量随着分割阈值T的增大而增大。当分割阈值T从孔隙结构的灰度级到达表面土骨架灰度级时，由于表面土骨架结构占大多数，故图像中每个像素的灰度级i所对应的总像素数量n i会存在一个突变，因此，找到这个从孔隙结构到表面土骨架结构突变的这个点对应的灰度级i，即确定分割阈值T的大小。
步骤SA4：获取灰度直方图曲线的二阶导数，具体过程如下：
步骤SA4.1：获取灰度直方图曲线的一阶导数，一阶导数为:
Figure PCTCN2019086147-appb-000009
其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量。
步骤SA4.2：获取灰度直方图曲线的二阶导数，二阶导数为：
Figure PCTCN2019086147-appb-000010
其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量。
步骤SA5：根据灰度直方图曲线的二阶导数a i和分割阈值T的取值范围，确定分割 阈值T的大小，具体过程如下：
步骤SA5.1：根据步骤SA4.2中的各个a i，拟合获取二阶导数曲线，参考图11，图11为膨润土的灰度直方图曲线的二阶导数曲线。
步骤SA5.2：确定在分割阈值T取值范围内，二阶导数曲线中的最大值a imax，同时确定最大值a imax所对应像素的灰度级i T,分割阈值T为：
T＝i T
其中，i T为二阶导数最大值所对应像素的灰度级。
参考图12，图12为膨润土最终的二值化图像。其中分割阈值T＝i T，故在膨润土的灰度直方图曲线中，灰度级i小于分割阈值T处所对应的范围为膨润土的孔隙结构，灰度级i大于分割阈值T处所对应的范围为膨润土的表面土骨架结构。
实施例2 含裂缝结构的岩土体材料
本实施例提供了一种岩土体材料数字图像的分割阈值确定方法，在本实施例中，具体地讲，待测岩土体材料选择压裂后的煤样。其中压裂后煤样的数字图像的分割阈值确定方法，具体如下：
步骤SB1：参考图8，图8为压裂后煤样的SEM图像，通过MATALB代码读取压裂后煤样的SEM图像，获取压裂后煤样SEM图像中每个像素的灰度级i以及每个像素的灰度级i所对应的总像素数量n i。
步骤SB2：获取压裂后煤样SEM图像的灰度直方图曲线，具体过程如下：
步骤SB2.1：根据每个像素的灰度级i、每个像素的灰度级i所对应的总像素数量n i以及如下公式，获取压裂后煤样灰度图像的灰度直方图曲线上的各个点：
Figure PCTCN2019086147-appb-000011
其中，i表示灰度级，N表示图像像素总数，n i表示图像中所有灰度级为i的像素的总像素个数，L表示灰度级的种类数。
步骤SB2.2：根据步骤SB2.1中的各个P(i)，拟合获取压裂后煤样灰度图像的灰度直方图曲线。参考图9，图9为压裂后煤样SEM图像的灰度直方图曲线。
步骤SB3：根据灰度直方图曲线，确定分割阈值T的取值范围，具体过程如下：
步骤SB3.1：根据灰度直方图曲线，确定灰度直方图曲线中的峰值数目，在本实施例中，参考图9可知，灰度直方图曲线中的峰值有两个，因此压裂后煤样的结构为裂缝结构。
步骤SB3.2：根据步骤SB3.1可知，压裂后煤样的灰度直方图曲线中有两个峰值，因此在本实施例中，需要确定两个峰值分别对应像素的灰度级i max1和i max2，其中，i max1为第一个峰值对应像素的灰度级，i max2为第二个峰值对应像素的灰度级。
步骤SB3.3：根据压裂后煤样的结构、相应峰值对应像素的灰度级i max1和i max2，确定分割阈值T的取值范围，因为压裂后煤样的结构为孔隙结构，且其灰度直方图曲线有两个峰值，故压裂后煤样数字图像的分割阈值T的取值范围为：
i max1<T<i max2
其中，i max1为第一个峰值对应像素的灰度级，i max2为第二个峰值对应像素的灰度级。
参考图9，含裂缝结构的岩土体材料的灰度直方图曲线均为双峰曲线，故压裂后煤样的结构为含裂缝结构的岩土体材料，在灰度级i从小到大变化的过程中，图像中第一个峰值处的像素代表着裂缝处的像素，而第二个峰值处的像素代表着表面骨架处的像素，因此含裂缝结构的岩土体材料的分割阈值T介于两个峰值之间。
步骤SB4：获取灰度直方图曲线的二阶导数，具体过程如下：
步骤SB4.1：获取灰度直方图曲线的一阶导数，一阶导数为:
Figure PCTCN2019086147-appb-000012
其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量。
步骤SB4.2：获取灰度直方图曲线的二阶导数，二阶导数为：
Figure PCTCN2019086147-appb-000013
其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量。
步骤SB5：根据灰度直方图曲线的二阶导数a i和分割阈值T的取值范围，确定分割阈值T的大小，具体过程如下：
步骤SB5.1：根据步骤SB4.2中的各个a i，拟合获取二阶导数曲线，参考图13，图13为压裂后煤样的灰度直方图曲线的二阶导数曲线。
步骤SB5.2：确定在分割阈值T取值范围内，二阶导数曲线中的最大值a imax，同时确定最大值a imax所对应像素的灰度级i T,分割阈值T为：
T＝i T
其中，i T为二阶导数最大值所对应像素的灰度级。
参考图14，图14为压裂后煤样最终的二值化图像。其中分割阈值T＝i T，故在压裂后煤样的灰度直方图曲线中，灰度级i小于分割阈值T处所对应的范围为压裂后煤样的裂缝结构，而灰度级i大于分割阈值T处所对应的范围则为压裂后煤样的表面土骨架结构。
以上示意性的对本发明及其实施方式进行了描述，该描述没有限制性，附图中所示的也只是本发明的实施方式之一，实际的结构和方法并不局限于此。所以，如果本领域的普通技术人员受其启示，在不脱离本发明创造宗旨的情况下，不经创造性的设计出与该技术方案相似的结构方式及实施例，均属于本发明的保护范围。

Claims (8)

  1. 一种岩土体材料数字图像的分割阈值确定方法，其特征在于，所述方法包括如下步骤：
    S1：获取岩土体材料的SEM图像的灰度直方图曲线；
    S2：根据所述灰度直方图曲线，确定分割阈值T的取值范围；
    S3：获取所述灰度直方图曲线的二阶导数；
    S4：根据所述灰度直方图曲线的二阶导数和分割阈值T的取值范围，确定所述分割阈值T的大小。
  2. 根据权利要求1所述的一种岩土体材料数字图像的分割阈值确定方法，其特征在于，所述步骤S1获取图像的灰度直方图曲线之前还包括：读取所述岩土体材料的SEM图像，获取所述SEM图像中每个像素的灰度级i、所述每个像素的灰度级i所对应的总像素数量n i。
  3. 根据权利要求1或2所述一种岩土体材料数字图像的分割阈值确定方法，其特征在于，所述步骤S1获取图像的灰度直方图曲线具体如下：
    S1.1：确定所述岩土体材料灰度图像中每个像素的灰度级i；
    S1.2：根据如下公式，获取所述岩土体材料灰度图像的灰度直方图曲线上各个点：
    Figure PCTCN2019086147-appb-100001
    其中，i表示灰度级，N表示图像像素总数，n i表示图像中所有灰度级为i的像素的总像素个数，L表示灰度级的种类数；
    S1.3：根据所述P(i)，拟合获取岩土体材料灰度图像的灰度直方图曲线。
  4. 根据权利要求3所述的一种岩土体材料数字图像的分割阈值确定方法，其特征在于，所述步骤S2确定分割阈值T的取值范围具体如下：
    步骤S2.1：根据所述灰度直方图曲线，确定所述灰度直方图曲线中的峰值数目；
    步骤S2.2：根据所述峰值数目，确定所述岩土体材料的结构；
    步骤S2.3：获取所述峰值对应像素的灰度级i max；
    步骤S2.4：根据所述岩土体材料的结构和峰值对应像素的灰度级i max，确定所述分割阈值T的取值范围。
  5. 根据权利要求4所述的一种岩土体材料数字图像的分割阈值确定方法，其特征在于，步骤S2.2确定所述岩土体材料的结构具体如下：
    如果所述灰度直方图曲线只有一个峰值，所述岩土体材料的结构为含孔隙结构；
    如果所述灰度直方图曲线有两个峰值，所述岩土体材料的结构为含裂缝结构。
  6. 根据权利要求5所述的一种岩土体材料数字图像的分割阈值确定方法，其特征在于，步骤S3获取所述灰度直方图曲线的二阶导数之前包括：获取所述灰度直方图曲 线的一阶导数，所述一阶导数为:
    Figure PCTCN2019086147-appb-100002
    其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量。
  7. 根据权利要求6所述的一种岩土体材料数字图像的分割阈值确定方法，其特征在于，确定所述孔隙结构的分割阈值T具体如下：
    SA4.1：所述灰度直方图曲线只有一个峰值，所述分割阈值T的取值范围为：
    T<i max
    其中，i max为峰值对应像素的灰度级；
    SA4.2：获取所述灰度直方图曲线的二阶导数，所述二阶导数为：
    Figure PCTCN2019086147-appb-100003
    其中，i表示灰度级，n i表示图像中每个像素的灰度级i所对应的总像素数量；
    SA4.3：根据所述二阶导数，确定在分割阈值T的取值范围内二阶导数的最大值a imax；
    SA4.4：确定所述二阶导数最大值a imax所对应像素的灰度级i T,所述分割阈值T为：
    T＝i T
    其中，i T为二阶导数最大值a imax所对应像素的灰度级。
  8. 根据权利要求6所述的一种岩土体材料数字图像的分割阈值确定方法，其特征在于，确定所述裂缝结构的分割阈值T具体如下：
    SB4.1：所述灰度直方图曲线有两个峰值，所述分割阈值T的取值范围为：
    i max1<T<i max2
    其中，i max1为第一个峰值对应像素的灰度级，i max2为第二个峰值对应像素的灰度级；
    SB4.2：获取所述灰度直方图曲线的二阶导数，所述二阶导数为：
    Figure PCTCN2019086147-appb-100004
    其中，n i表示图像中所有灰度级为i的像素的总像素个数，i表示灰度级；
    SB4.3：根据所述二阶导数，确定在分割阈值T的取值范围内二阶导数的最大值a imax；
    SB4.4：确定所述二阶导数最大值a imax所对应像素的灰度级i T,所述分割阈值T为：
    T＝i T
    其中，i T为二阶导数最大值a imax所对应像素的灰度级。
PCT/CN2019/086147 2019-03-05 2019-05-09 一种岩土体材料数字图像的分割阈值确定方法 WO2020177215A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/965,307 US20210201497A1 (en) 2019-03-05 2019-05-09 Method for determining segmentation threshold of digital image of rock-soil material

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910174541.9A CN110021030B (zh) 2019-03-05 2019-03-05 一种岩土体材料数字图像的分割阈值确定方法
CN201910174541.9 2019-03-05

Publications (1)

Publication Number Publication Date
WO2020177215A1 true WO2020177215A1 (zh) 2020-09-10

Family

ID=67189412

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/086147 WO2020177215A1 (zh) 2019-03-05 2019-05-09 一种岩土体材料数字图像的分割阈值确定方法

Country Status (3)

Country Link
US (1) US20210201497A1 (zh)
CN (1) CN110021030B (zh)
WO (1) WO2020177215A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113702258A (zh) * 2021-08-19 2021-11-26 国家烟草质量监督检验中心 一种卷烟轴向孔隙分布的检测方法
CN116433663A (zh) * 2023-06-13 2023-07-14 肥城恒丰塑业有限公司 一种土工格室质量智能检测方法
CN116503394A (zh) * 2023-06-26 2023-07-28 济南奥盛包装科技有限公司 基于图像的印刷制品表面粗糙度检测方法

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110458816B (zh) * 2019-08-06 2021-04-27 北京工商大学 一种基于阈值回归的纤维材料孔隙率分析方法
CN112085693B (zh) * 2020-06-24 2022-09-20 中国科学院武汉岩土力学研究所 土石混合体内部结构的孔隙比评估及形态重建方法及系统
CN113112453B (zh) * 2021-03-22 2022-03-22 深圳市华启生物科技有限公司 胶体金检测卡识别方法、系统、电子设备及存储介质
CN113702259A (zh) * 2021-08-19 2021-11-26 国家烟草质量监督检验中心 一种卷烟整体孔隙均匀度的检测方法
CN114168884B (zh) * 2021-12-02 2024-08-02 成都工业学院 一种岩土样本的概率分布获取方法
CN114705606B (zh) * 2022-04-06 2024-07-02 同济大学 一种基于网络化分析的岩石内部关键渗流节点的封堵方法
CN115049566B (zh) * 2022-08-15 2022-10-25 聊城扬帆田一机械有限公司 一种平板夯激振模式智能调节系统
CN116977230B (zh) * 2023-09-22 2024-01-02 济宁市质量计量检验检测研究院(济宁半导体及显示产品质量监督检验中心、济宁市纤维质量监测中心) 一种扫描电子显微镜图像优化增强方法
CN117291945B (zh) * 2023-11-24 2024-02-13 山东省济宁生态环境监测中心(山东省南四湖东平湖流域生态环境监测中心) 基于图像数据的土壤腐蚀污染检测预警方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101710425A (zh) * 2009-12-25 2010-05-19 南京航空航天大学 基于图像灰度梯度和灰度统计直方图的自适应预分割方法
CN105654501A (zh) * 2016-02-22 2016-06-08 北方工业大学 基于模糊阈值的自适应图像分割方法
CN106340029A (zh) * 2016-08-23 2017-01-18 湖南文理学院 基于Beta‑Gamma散度的灰度图像阈值分割方法
CN107590815A (zh) * 2017-09-07 2018-01-16 陕西师范大学 基于萤火虫群优化法的图像多阈值分割方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101710425A (zh) * 2009-12-25 2010-05-19 南京航空航天大学 基于图像灰度梯度和灰度统计直方图的自适应预分割方法
CN105654501A (zh) * 2016-02-22 2016-06-08 北方工业大学 基于模糊阈值的自适应图像分割方法
CN106340029A (zh) * 2016-08-23 2017-01-18 湖南文理学院 基于Beta‑Gamma散度的灰度图像阈值分割方法
CN107590815A (zh) * 2017-09-07 2018-01-16 陕西师范大学 基于萤火虫群优化法的图像多阈值分割方法

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113702258A (zh) * 2021-08-19 2021-11-26 国家烟草质量监督检验中心 一种卷烟轴向孔隙分布的检测方法
CN113702258B (zh) * 2021-08-19 2024-01-19 国家烟草质量监督检验中心 一种卷烟轴向孔隙分布的检测方法
CN116433663A (zh) * 2023-06-13 2023-07-14 肥城恒丰塑业有限公司 一种土工格室质量智能检测方法
CN116433663B (zh) * 2023-06-13 2023-08-18 肥城恒丰塑业有限公司 一种土工格室质量智能检测方法
CN116503394A (zh) * 2023-06-26 2023-07-28 济南奥盛包装科技有限公司 基于图像的印刷制品表面粗糙度检测方法
CN116503394B (zh) * 2023-06-26 2023-09-08 济南奥盛包装科技有限公司 基于图像的印刷制品表面粗糙度检测方法

Also Published As

Publication number Publication date
CN110021030B (zh) 2023-04-25
CN110021030A (zh) 2019-07-16
US20210201497A1 (en) 2021-07-01

Similar Documents

Publication Publication Date Title
WO2020177215A1 (zh) 一种岩土体材料数字图像的分割阈值确定方法
US8909508B2 (en) Petrographic image analysis for determining capillary pressure in porous media
CN105445160B (zh) 一种沥青混合料的空隙特征及其提取方法
Liu et al. A new method for threshold determination of gray image
CN110443793A (zh) 一种沥青混合料空隙分布均匀性评价方法
CN108061697B (zh) 土体三维孔隙率计算方法
CN102841220B (zh) 一种基于孔隙率的黏土扫描电镜照片图像分割方法
CN113158490A (zh) 致密砂岩储层渗透率计算模型的建立方法
CN104268830A (zh) 基于数字图像确定不均匀岩土材料渗透系数的方法
Ren et al. Characterization of internal pore size distribution and interconnectivity for asphalt concrete with various porosity using 3D CT scanning images
CN116542108A (zh) 一种基于神经网络学习的道路水稳基层开裂性能预测方法
CN110163826A (zh) 一种黄土微结构图像的处理方法
CN109406364A (zh) 一种纤维过滤介质结构孔隙率的测算方法
Kozłowski et al. Application of SEM to analysis of permeability coefficient of cohesive soils
CN104217424A (zh) 冰湖终碛垄颗粒的分析方法
CN110988004B (zh) 一种粒间压溶成因石英胶结物含量的评价方法
CN113269712B (zh) 一种基于铸体薄片构建储层岩石毛管压力曲线的方法
Moraru et al. Euler number: a method for statistical analysis of ancient pottery porosity
Ahamad et al. Digital analysis of geo-referenced concrete scanning electron microscope (SEM) images
CN114777709B (zh) 一种基于子体块分离的dvc微裂纹表征方法
Yiediboe et al. The Relationship between the Physical Properties of Soil and Shape Factors of its Fragmented Aggregates: A Two-Dimensional Digital Image Processing and Analysis Approach
CN117744532B (zh) 一种辨识低渗煤体瓦斯扩散-渗流有效边界的方法
CN117129509B (zh) 一种基于ki-ct计算页岩裂缝核磁共振测井t2截止值的方法
Tian et al. 3d damage identification of soil rock mixture based on image processing technology
Bolshakov et al. Investigation of the pore space structure by a scanning electron microscope using the computer program collector

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19918274

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19918274

Country of ref document: EP

Kind code of ref document: A1